Biostats (United States)

Anterior shoulder instability, which includes dislocations and subluxations, is a significant concern in pediatric populations (≤19 years). Following an initial dislocation or subluxation, these patients often develop recurrent instability, which can result in long-term functional compromise. While timely and appropriate treatment is crucial, there remains a lack of consensus in the literature regarding the optimal management of pediatric shoulder instability. This systematic review aims to compare the efficacy of conservative and operative treatments for traumatic, isolated anterior shoulder instability in pediatric populations using recurrence of instability and return to play (RTP) at pre-injury levels as outcome measures. This study also provides a contemporary analysis that reflects evolving treatment strategies.This systematic review targeted studies published between 2013 and 2023 that evaluated pediatric patients (≤19 years), most between 13 and 19 years of age, who received conservative or operative treatment for a first-time or recurrent traumatic isolated anterior shoulder dislocation or subluxation. Statistical analysis was performed to compare rates of recurrence of instability and RTP within each group. Pooled effects (odds ratios) were estimated using fixed-effects models where heterogeneity was absent; otherwise, random-effects models were applied.A total of 1,459 patients (1,468 shoulders) met the inclusion criteria. Of these, 593 (40.4%) underwent conservative treatment, while 885 (60.3%) received operative interventions. Recurrence rates of instability were higher in the conservative group (236/543, 43.5%) compared to the primary operative group (182/875, 20.8%) and secondary operative group (2/26, 7.7%). RTP rates included 145/193 (75.1%) of conservatively treated patients, 380/480 (79.2%) of primary operative patients, and 8/11 (72.7%) of secondary operative patients returning to pre-injury levels. Among the four studies that directly compared conservative and primary operative treatments, the nonoperative group was more likely to have recurrence compared to the primary operative group (OR = 6.90; 95% CI, 2.28-20.91; p < 0.001). One of these studies was excluded due to methodological differences. A subsequent meta-analysis revealed a significantly higher likelihood of recurrent instability in conservatively treated patients (odds ratio (OR) = 9.55; 95% confidence interval (CI): 5.10-17.88; p < 0.001). In contrast, there was no statistically significant difference in RTP between groups (OR = 3.11; 95% CI: 0.31-30.97; p = 0.33).The findings support early surgical intervention in pediatric patients to reduce recurrence and improve functional outcomes. Conservative management, while recently shown to be successful for patients with less severe injuries such as subluxations, is still primarily associated with higher recurrence rates compared to operative treatment. Further studies are needed to refine treatment protocols by distinguishing effective strategies for subluxations versus dislocations. Future research should also explore the influence of factors such as sex, skeletal maturity, and activity level in determining optimal management strategies in pediatric patients.

Abstract We introduce GeneJepa, a self-supervised foundation model that learns a predictive world model of single-cell transcriptomes. Based on the Joint-Embedding Predictive Architecture, GeneJepa predicts latent representations of masked gene sets from visible context, a shift away from reconstructing noisy expression values and toward world-model style inference over cellular state. To realize this at scale, a Perceiver encoder handles variable gene sets at fixed cost, and a tokenizer jointly represents gene identity and continuous expression using Fourier features. Trained on the Tahoe-100M atlas, GeneJepa learns general representations that transfer across tissues and datasets. On downstream tasks, including drug response and perturbation prediction, it surpasses strong baselines and enables test-time scaling by progressively enlarging the cross-attention over the gene set, trading a small read cost for higher accuracy at inference. GeneJepa moves toward foundation models that reason over gene–gene relations, enabling applications in annotation, prediction, and in-silico discovery.

SummaryBackgroundKnowledge of the contribution of the pathogen to the heterogeneity of the host response to infection is limited. We aimed to compare the host response in critically ill patients with a bloodstream infection (BSI).MethodsRNA profiles were determined in blood obtained between one day before and after a positive blood culture. Differential expression and pathway analyses were performed on independent patients’ samples by RNA sequencing (discovery) or microarray (validation). Additional patients were included for the discovery and validation of transcriptome classifiers of pathogen-specific BSIs. Twenty biomarkers reflecting key host response pathways were measured in blood.FindingsWe included 341 patients, among which 255 with BSI, 25 with viral infection and 61 non-infectious controls. The cultured pathogen explained 41·8% of the blood transcriptomic variance in patients with BSI. Gene set enrichment analysis showed a global resemblance between monomicrobial BSIs caused by Streptococcus, Staphylococcus aureus and Escherichia coli, which were clearly different from BSI caused by coagulase-negative staphylococci or Enterococcus. BSI by Streptococcus was associated with the highest number of differentially expressed genes, indicating strong innate and adaptive immune activation. An eight-gene streptococcal classifier performed well across different Streptococcus species, and was validated in external cohorts. Plasma biomarker profiling showed that E. coli BSI was associated with the strongest response in the cytokine and systemic inflammation domain, and S. aureus BSI with the strongest endothelial cell activation.InterpretationThe causative pathogen explains a substantial part of the heterogeneity of the host response in critically ill patients with BSI.Funding10.13039/501100006020Center for Translational Molecular Medicine and the 10.13039/501100000780European Commission.

Abstract Rationale: Children with bronchopulmonary dysplasia (BPD) are at increased risk of developing asthma and poor lung function. Little is known about social drivers of health, particularly neighborhood resources, and long-term BPD outcomes. We aimed to evaluate the link between neighborhood social risk at birth, using the Childhood Opportunity Index (COI) 3.0, and asthma prevalence among a cohort of children with BPD. Methods: Children aged 6-12 years with BPD were enrolled in the Air Quality, Environment and Respiratory Outcomes in BPD (AERO-BPD; NCT04107701) study. Geocoded birth addresses were linked to COI levels (very low, low, moderate, high, and very high opportunity) using established methodology. Patient questionnaires assessed sociodemographic and medical history data, including the primary outcome of asthma diagnosis. Neonatal medical records provided gestational age (GA), birth weight (BW), and respiratory support at 36 weeks corrected GA. Overall state ranked COI and COI across 3 domains – education, health and environment, and social and economic – were reported. Data was summarized using descriptive statistics. To compare measures across COI levels, we used the Chi-Square or Fisher's exact test for categorical variables and the Kruskal-Wallis test for continuous variables. Results: Our cohort consisted of 157 children (55.4% male, median age: 8 years [IQR 7-10], mean (sd) GA: 27 (2) weeks, mean (sd) BW: 919 (290) grams). At 36 weeks, 25.5% were in room air. In the first year of life, 24.2% of participants lived in very low, 17.2% in low, 15.3% in moderate, 22.3% in high, and 15.9% in very high COI neighborhoods. BW varied significantly across COI levels (p = 0.04), with BW notably lower in the very low as compared to very high COI neighborhoods (822 grams v. 976 grams). A statistically nonsignificant trend towards higher asthma prevalence was seen with lower COI (very low COI, 55.3%; low COI, 48.2%; high COI, 40.0%, very high COI, 40.0%; p = 0.42). COI level was not associated with GA (p = 0.10) or ability to wean to room air by 36 weeks corrected (p = 0.76). Conclusion: In the AERO-BPD cohort, COI levels were evenly distributed across COI categories. Those in lower opportunity neighborhoods were more likely to have lower BW, despite no differences in GA. Although not statistically significant, the prevalence of school-age asthma was highest among the lowest COI levels. Further analysis is needed to determine the effect of social and environmental determinants of health on outcomes in BPD.

Abstract Generative pre-trained models, such as BERT and GPT, have demonstrated remarkable success in natural language processing and computer vision. Leveraging the combination of large-scale, diverse datasets, transformers, and unsupervised learning, these models have emerged as a promising method for understanding complex systems like language. Despite the apparent differences, human language and biological systems share numerous parallels. Biology, like language, is a dynamic, interconnected network where biomolecules interact to create living entities akin to words forming coherent narratives. Inspired by this analogy, we explored the potential of using transformer-based unsupervised model development for analyzing biological systems and proposed a framework that can ingest vast amounts of biological data to create a foundational model of biology using BERT or GPT. This framework focuses on the concept of a ‘biostate,’ defined as a high-dimensional vector encompassing various biological markers such as genomic, proteomic, transcriptomic, physiological, and phenotypical data. We applied this technique to a small dataset of single-cell transcriptomics to demonstrate its ability to capture meaningful biological insights into genes and cells, even without any pre-training. Furthermore, the model can be readily used for gene network inference and genetic perturbation prediction.

Mantle cell lymphoma (MCL) is a rare, incurable lymphoma subtype characterized by heterogeneous outcomes. To better understand the clinical behavior and response to treatment, predictive biomarkers are needed. Using residual archived material from patients enrolled in the MCL3001 (RAY) study, we performed detailed analyses of gene expression and targeted genetic sequencing. This phase III clinical trial randomized patients with relapsed or refractory MCL to treatment with either ibrutinib or temsirolimus. We confirmed the prognostic capability of the gene expression proliferation assay MCL35 in this cohort treated with novel agents; it outperformed the simplified MCL International Prognostic Index in discriminating patients with different outcomes. Regardless of treatment arm, our data demonstrated that this assay captures the risk conferred by known biological factors, including increased MYC expression, blastoid morphology, aberrations of TP53, and truncated CCND1 3' untranslated region. We showed the negative impact of BIRC3 mutations/deletions on outcomes in this cohort and identified that deletion of chromosome 8p23.3 also negatively impacts survival. Restricted to patients with deletions/alterations in TP53, ibrutinib appeared to abrogate the deleterious impact on outcome. These data illustrate the potential to perform a molecular analysis of predictive biomarkers on routine patient samples that can meaningfully inform clinical practice.

Electroanatomic mapping systems are increasingly used during ablations to decrease the need for fluoroscopy and therefore radiation exposure. For left-sided arrhythmias, transseptal puncture is a common procedure performed to gain access to the left side of the heart. We aimed to demonstrate the radiation exposure associated with transseptal puncture. Data were retrospectively collected from the Catheter Ablation with Reduction or Elimination of Fluoroscopy registry. Patients with left-sided accessory pathway-mediated tachycardia, with a structurally normal heart, who had a transseptal puncture, and were under 22 years of age were included. Those with previous ablations, concurrent diagnostic or interventional catheterisation, and missing data for fluoroscopy use or procedural outcomes were excluded. Patients with a patent foramen ovale who did not have a transseptal puncture were selected as the control group using the same criteria. Procedural outcomes were compared between the two groups. There were 284 patients in the transseptal puncture group and 70 in the patent foramen ovale group. The transseptal puncture group had a significantly higher mean procedure time (158.8 versus 131.4 minutes, p = 0.002), rate of fluoroscopy use (38% versus 7%, p < 0.001), and mean fluoroscopy time (2.4 versus 0.6 minutes, p < 0.001). The acute success and complication rates were similar. Performing transseptal puncture remains a common reason to utilise fluoroscopy in the era of non-fluoroscopic ablation. Better tools are needed to make non-fluoroscopic transseptal puncture more feasible.